A sensitivity analysis of a hull's local characteristic parameters on ship resistance performance

作者：Zhang, Heng^*; Liu, Zuyuan; Zhan, Chengsheng; Feng, Baiwei

来源：Journal of Marine Science and Technology (Japan), 2016, 21(4): 592-600.

DOI：10.1007/s00773-016-0378-x

• Sensitivity analysis
• Dimensionality reduction
• RBF-based neural network
• Partial least squares regression
• Uniform design
• Hull form optimization

Simulation-based design hull form optimization has become one of the most effective ways to develop energy-saving ships. However, the traditional method is time consuming and inefficient, so it is difficult to apply it to practical engineering. One of the reasons for this difficulty is that there are numerous parameters for expressing the hull form, making it difficult to find the optimal solution within a short amount of time. Therefore, this paper presents a sensitivity analysis method based on radial basis function and partial least squares regression neural network with uniform design for sampling, to study the sensitivity of local hull parameters of a class of container ships on wave-making resistance performance, and to reduce the number of dimensions in the design space. According to the results of sensitivity analysis, the parameters that are of negligible influence on the objective function can be removed from the optimization model. Comparison of results of the original model and the simplified model shows that the presented sensitivity analysis method can effectively reduce the dimension of the design space so that the time consumed to achieve the optimum is decreased.

• 出版日期2016-12
• 单位武汉理工大学